Custom fieldname support

Add support for validating CSVs that don't include a header row. The
implementation allows a user to define a sequence of field names and
passes these through to the underlying `DictReader` instances. The
`fieldnames` attribute is `None` by default, which retains the existing
behavior of inferring field names from a header row in the source CSV.

Resolves di#88.

README.rst

@@ -343,6 +343,13 @@ Running Vlads Programatically
       List of validators. Optional, defaults to the class variable `validators`
       if set, otherwise uses `EmptyValidator` for all fields.
 
+  :``fieldnames=None``:
+      Sequence of field names to be passed through to the underlying
+      `csv.DictReader` instance. If provided, the reader will use these field
+      names instead of inferring them from the input CSV's first row. Intended
+      only for use with CSVs that do not have header rows. Optional, defaults
+      to `None`.
+
   :``delimiter=','``:
       The delimiter used within your csv source. Optional, defaults to `,`.
 

tests/test_vlads.py

@@ -90,6 +90,32 @@ class TestVlad(Vlad):
     assert not TestVlad(source=source).validate()
 
 
+def test_explicit_fieldnames():
+    source = String("Dracula,Vampire")
+
+    class TestVlad(Vlad):
+        validators = {
+            "Name": [UniqueValidator()],
+            "Status": [SetValidator(["Vampire", "Not A Vampire"])],
+        }
+        fieldnames = ["Name", "Status"]
+
+    assert TestVlad(source=source).validate()
+
+
+def test_explicit_fieldnames_conflict_fails():
+    source = LocalFile("vladiate/examples/vampires.csv")
+
+    class TestVlad(Vlad):
+        validators = {
+            "Name": [UniqueValidator()],
+            "Status": [SetValidator(["Vampire", "Not A Vampire"])],
+        }
+        fieldnames = ["Name", "Status"]
+
+    assert not TestVlad(source=source).validate()
+
+
 def test_fails_validation():
     source = LocalFile("vladiate/examples/vampires.csv")
 

vladiate/vlad.py

@@ -17,6 +17,7 @@ def __init__(
         file_validation_failure_threshold=None,
         quiet=False,
         row_validators=[],
+        fieldnames=None,
     ):
         self.logger = logs.logger
         self.failures = defaultdict(lambda: defaultdict(list))
@@ -26,6 +27,7 @@ def __init__(
         self.source = source
         self.validators = validators or getattr(self, "validators", {})
         self.row_validators = row_validators or getattr(self, "row_validators", [])
+        self.fieldnames = fieldnames or getattr(self, "fieldnames", None)
         self.delimiter = delimiter or getattr(self, "delimiter", ",")
         self.line_count = 0
         self.ignore_missing_validators = ignore_missing_validators
@@ -124,15 +126,19 @@ def _log_missing(self, missing_items):
         )
 
     def _get_total_lines(self):
-        reader = csv.DictReader(self.source.open(), delimiter=self.delimiter)
+        reader = csv.DictReader(
+            self.source.open(), delimiter=self.delimiter, fieldnames=self.fieldnames
+        )
         self.total_lines = sum(1 for _ in reader)
         return self.total_lines
 
     def validate(self):
         self.logger.info(
             "\nValidating {}(source={})".format(self.__class__.__name__, self.source)
         )
-        reader = csv.DictReader(self.source.open(), delimiter=self.delimiter)
+        reader = csv.DictReader(
+            self.source.open(), delimiter=self.delimiter, fieldnames=self.fieldnames
+        )
 
         if not reader.fieldnames:
             self.logger.info(